Forum Discussion
Data Masked Column Giving Wrong Records In Visual
Hi All,
In my report I have a slicers like employersname and Members slicers along with date range slicer. Members slicer will show the coressponding members for the selected Employersname slicer.
I created a measure with the attached logic for my required columns data masking. The issue is when I drop the firstname masked cloumn which I created, that is not filtering the visaul data as per the employername and Members slicers values selection. For example for X employer there are 2 Members. But when I place the newly created Firstname masked column, it pulling all the records from the backend with out considering the slicers passed values. it seems like mentioned logic is applying masking on every record by scanning the entire table.
Can you please suggest how to proceed in this scenario to show the visual data accroding to slicer values selections with data masking for the required fileds in the layout.
Thanks,
Sri.
Hi Koritala ,
Awesome, thanks for confirming you’re working with a single view and no joins that definitely keeps things simple. In the example DataNinja777 shared, they just used a static email to show how the masking works, but you can totally make it dynamic with the USERPRINCIPALNAME function so it updates automatically for whoever’s logged in.
Here’s how you can set up your Can See PII measure:
Can See PII = IF( USERPRINCIPALNAME() IN { "[email protected]", "[email protected]" }, TRUE(), FALSE() )
This checks if the signed-in user is one of the people allowed to view unmasked data. Just swap out those emails for the real ones in your org. Once you’ve got that, you can plug it into your masking logic like you did in your test:
FirstName Masked = VAR vFirstName = MAXX( VALUES( 'YourView'[FirstName] ), 'YourView'[FirstName] ) RETURN IF( [Can See PII], vFirstName, IF( NOT ISBLANK(vFirstName), REPT("*", LEN(vFirstName)), BLANK() ) )
Since you’re only using one view, your slicers will filter everything automatically, so you don’t need to mess with relationships or TREATAS. The masking will just work with your slicers and whoever’s logged in.
If you want more details, check out the Microsoft docs:
USERPRINCIPALNAME – https://learn.microsoft.com/en-us/dax/userprincipalname-function-dax
SELECTEDVALUE – https://learn.microsoft.com/en-us/dax/selectedvalue-function-dax
TREATAS – https://learn.microsoft.com/en-us/dax/treatas-function-daxBest Regards,
Tejaswi.
Community Support
8 Replies
hi Koritala
it appears that your measure dont have a proper filter context relationship to your slicers, you can try using calculated column for this purpose, otherwise you have to apply filter context in your measure properly.
My Work:
Analytic Pulse Blog
Sharing data & BI insights
Docynx Productivity Tools
Tools to simplify workflowHi,
Thanks for your response.
We can't create as calculated column insteated of measure as we are using userprincipalname function in logic.
Thanks,
Sri
Hi Koritala ,
It appears that your filtering is not working in the intended way. I have created a pbix file to demonstrate the case.
FirstName Masked (Broken) = VAR vFirstName = CALCULATE ( MAXX ( VALUES ( 'Members'[FirstName] ), 'Members'[FirstName] ), ALL ( 'Members' ) ) RETURN IF ( NOT ISBLANK ( vFirstName ), REPT ( "*", LEN ( vFirstName ) ), BLANK () )Filtering will work properly by using the following measure:
FirstName Masked (Working) = VAR vFirstName = MAXX ( VALUES ( 'Members'[FirstName] ), 'Members'[FirstName] ) RETURN IF ( [Can See PII], vFirstName, IF ( NOT ISBLANK ( vFirstName ), REPT ("*", LEN ( vFirstName ) ), BLANK () ) )When the filtering is fixed, the crossjoining of the masking will disappear as shown above.
I am attaching the pbix file for your reference.
Best regards,
Hi DataNinja777,
Thanks for sharing pbxi.
One thing I want to tell you that in my case, I have only one view where all my report columns are coming. There are no multiple tables or joins.
Also, can you please confirm me, you have made static user id { "[email protected]" }. In my case how to write the expression exactly?
Thanks,
Sri.
Hi DataNinja777,
In continoution to my previous message, I have a users table with columns like "Email", "PIIUser(Y/N)".
In this case, if I wan to follow your below user validation, Can you send me the exact logic please? Because it should be dynamically work based on user login id.Please let me know if need any further details.
Hi Koritala ,
Awesome, thanks for confirming you’re working with a single view and no joins that definitely keeps things simple. In the example DataNinja777 shared, they just used a static email to show how the masking works, but you can totally make it dynamic with the USERPRINCIPALNAME function so it updates automatically for whoever’s logged in.
Here’s how you can set up your Can See PII measure:
Can See PII = IF( USERPRINCIPALNAME() IN { "[email protected]", "[email protected]" }, TRUE(), FALSE() )
This checks if the signed-in user is one of the people allowed to view unmasked data. Just swap out those emails for the real ones in your org. Once you’ve got that, you can plug it into your masking logic like you did in your test:
FirstName Masked = VAR vFirstName = MAXX( VALUES( 'YourView'[FirstName] ), 'YourView'[FirstName] ) RETURN IF( [Can See PII], vFirstName, IF( NOT ISBLANK(vFirstName), REPT("*", LEN(vFirstName)), BLANK() ) )
Since you’re only using one view, your slicers will filter everything automatically, so you don’t need to mess with relationships or TREATAS. The masking will just work with your slicers and whoever’s logged in.
If you want more details, check out the Microsoft docs:
USERPRINCIPALNAME – https://learn.microsoft.com/en-us/dax/userprincipalname-function-dax
SELECTEDVALUE – https://learn.microsoft.com/en-us/dax/selectedvalue-function-dax
TREATAS – https://learn.microsoft.com/en-us/dax/treatas-function-daxBest Regards,
Tejaswi.
Community Support
